Eur. Phys. J.

C (2020) 80:862
https://doi.org/10.1140/epjc/s10052-020-08443-4

Regular Article - Theoretical Physics

Interacting dark matter and cosmic acceleration

Víctor H. Cárdenas1,a , Samuel Lepe2,b
1 Instituto de Física y Astronomía, Universidad de Valparaíso, Gran Bretaña 1111, Valparaiso, Chile
2 Instituto de Física, Pontificia Universidad Católica de Valparaíso, Avenida Brasil 2950, Valparaiso, Chile

Received: 8 July 2020 / Accepted: 5 September 2020

© The Author(s) 2020

Abstract We study the effect of an explicit interaction nent [4], somehow a dynamical cosmological constant, and
between two scalar fields components describing dark matter also models where the gravitational theory is modified [5].
in the context of a recent proposal framework for interaction. Among the tricks that have been proposed to alleviate the
We find that, even assuming a very small coupling, it is suf- cosmic coincidence problem is to assume that DE (what-
ficient to explain the observational effects of a cosmological ever it is) is coupled to dark matter (DM). This is appealing
constant, and also overcome the problems of the  CDM because both dark components are detected only by their
model without assuming an exotic dark energy. gravitational effects, and so they can be confused and is not
easy to discriminate each component [6]. The usual way to
express an interaction between DE and DM [7–13] is intro-
ducing a new term Q such that
1 Introduction

In the context of the standard model of cosmology, the sim- ρ̇m + 3H (ρm + pm ) = −Q , (1)
plest way we can describe the observations that type Ia super- ρ̇de + 3H (ρde + pde ) = Q , (2)
nova are dimmer than expected [1,2] is by introducing –
by hand – a cosmological constant, leading to the claimed where a dot denotes differentiation with respect to time, H
accelerated expansion and to establish the so far successful is the Hubble rate, ρm and ρde are the energy densities, pm
Lambda Cold Dark Matter (LCDM) model. and pde are the pressures. If Q > 0 DM releases energy
Although this model agreed with almost every observa- into DE, while for Q < 0 the energy flows in the opposite
tional test, from a theoretical point of view the model can not direction. In the literature it is usual to take the function Q
be taken seriously. First of all, assuming that this model is be proportional to the energy densities, as Q = 3H γρm for
valid requires us to accept that we live right in a very special example. Fixing Q enable us to completely solve the system
time in the history of the universe, something like (again) and find the solutions ρm (a) and ρde (a).
positioning in the center of the universe (this time including To the DE problem, we have to add the DM problem.
the temporal coordinate). It can be fortuitous, but then this Although a puzzle, the DM problem is of a different nature
should be consider a huge “cosmic coincidence”. Another compared to the DE one. We have particle candidates that
worried around the LCDM model is lambda itself, . What is are under search and there is a certain consensus that this
it? Why it has today this particular value of 1.19×10−52 m−2 can be solved in the next years. This component, although of
? Does  evolve with time? All these questions (and some a non-baryonic nature, is perfectly possible to exist, beyond
more) drives the quest for new ideas that replace  with the standard model of particle physics. There are no weird
something else, to describe what we observe but without the features (such as negative pressure, for example) that need
theoretical problems (or worries) we mentioned before. We to be invoked to hold a model.
have named this “something else” component, dark energy A huge effort has been made looking for evidence of non-
(DE) [3]. Proposals that try to shed light into this problem are baryonic DM through basically three ways: indirect detec-
those assuming the existence of a quintessence field compo- tion – when DM particles produce Standard Model particles
(photons, electron/positrons, neutrinos etc) – by direct detec-
tion methods – when Standard Model particles recoil from
a e-mail: [email protected] (corresponding author) collisions with invisible DM – and from colliders – where
b e-mail: [email protected] collisions of Standard Model particles may produce missing

0123456789().: V,-vol 123

 862 Page 2 of 11 Eur. Phys. J. C (2020) 80:862

energy or decay products [14]. Direct-detection experiments where LGR is the Einstein-Hilbert Lagrangian and the DM
rely on the scattering of dark-matter particles from the halo Lagrangians Lm and Lx are assumed to be of the type used in
of the Milky Way in a detector on Earth. The direct-detection the LCDM model, i.e. the Lagrangian of the free DM with no
rate depends on the local dark-matter density, currently esti- interaction. The interaction between both DM – the m and
mated to be ρ = 0.39 ± 0.03 GeV cm3 . Dark matter cannot the x components – is introduced as is well known in field
only be detected directly in dedicated experiments searching theory, through an explicit Lagrangian term Lint . From (3)
for nuclear recoils from the scattering of dark-matter parti- the field equations are
cles or produced in particle accelerators such as the LHC, but
it can also reveal its existence indirectly. The total number
of dark-matter particles does not change significantly after 3H 2 = ρm + ρx + ρint , (4)
freeze-out in the early universe, but their spatial distribution 2 Ḣ + 3H = − pm − px − pint ,
2
(5)
changes considerably during structure formation. The very ρ̇m + 3H (ρm + pm ) = Q m , (6)
selfannihilation that plays a central role in this freeze-out can
ρ̇x + 3H (ρx + px ) = Q x , (7)
give rise to a significant flux of −rays, neutrinos, and even
antimatter such as antiprotons and positrons, especially in
regions with large dark-matter density. Moreover the total energy conservation implies
For example, recently there is interest in exploring
the astrophysical consequences of an explicit interaction ρ̇int + 3H (ρint + pint ) + Q m + Q x = 0 , (8)
between DM (whatever it is) with radiation. In [15–17] the
authors find that even a small interaction is sufficient to alle- showing that the new functions are actually constrained. The
viate the small satellite problem, for example. system (4–7) was written before in [20]. In the system (4–
In this letter we want to explore the consequences of a 7) the Q i functions are obtained from the variation of Lint
small interaction between two DM components (none of respect to the ith degrees of freedom, so they are in gen-
them being exotic) at cosmological scale using a slight mod- eral different functions. In the next section we study some
ification to the interaction framework we have described examples solutions.
above. The analysis of the theoretical predictions are also com-
plemented with a statistical study. In this work we concen-
trate in using data from Type Ia supernova to constraint the
2 Coupling dark matter and analysis values of the parameters in the models. In particular we use
the type Ia supernova data from the Pantheon set [23]. This
If DM is a weakly interacting massive particle (WIMP), the sample consist in 1048 spectroscopically confirmed SNIa in
interactions are essential to obtain the relic abundance that the range 0.01 < z < 2.3. We compute the residuals μ − μth
support the whole scenario, the so called “WIMP miracle”. an minimize the quantity
Also the interactions between DM particles with those of the
standard model (SM) are essential to the extended campaigns χ 2 = (μ − μth )T C −1 (μ − μth ), (9)
of detection (direct and indirect) of DM using observatories
in the ground and in space. Further, as it was mentioned in where μth = 5 log10 (d L (z)/10 pc) gives the theoretical dis-
the previous section, there are increasing evidence for the tance modulus, d L (z) is the luminosity distance, C is the
astrophysical consequences of the interaction between DM covariance matrix released in [23], and the observational dis-
and photons, not only as can be seen in the power spectrum tance modulus takes the form
of the cosmic microwave background radiation [18,19], but
also in the context of the small scale problem such as the μ = m − M + α1 X − α2 Y, (10)
small satellite problem [15–17]. Then it is natural to study
their consequences at cosmological level. where m is the maximum apparent magnitude in band B,
Let us study the background cosmic evolution assuming an X is related to the widening of the light curves, and Y cor-
explicit interaction between two species of DM particles, one rects the color. usually, the cosmology – specified here by
indicated by an m subscript and other by x. No cosmological μth – is constrained along with the parameters M, α1 and
constant term or an exotic component is introduced. Gravity α2 . The analisys is performed using a public code known as
is described by general relativity (GR). On general grounds emcee [21]. This is a stable, well tested Python implemen-
the system can be described by a Lagrangian of the type tation of the affine-invariant ensemble sampler for Markov
chain Monte Carlo (MCMC) proposed by Goodman & Weare
Ltot = LGR + Lm + Lx + Lint , (3) [22].

123
Eur. Phys. J. C (2020) 80:862 Page 3 of 11 862

3 Some model examples

3.1 Symmetric model

Let us study first a simple model for the interaction between

two DM components that we shall denotate with subscripts m
and x. Let say Q m = −3H αρm and Q x = −3Hβρx , expres-
sions already used in the literature. Notice that these two fac-
tors do not need to be small, neither equals. We assume that
α and β are different from zero if some interaction operates
between these components.
From Eqs. (6) and (7) we obtain
ρm = ρm0 a −3(1+α) , ρx = ρx0 a −3(1+β) , (11) Fig. 1 Here we reconstruct both (14) (continuous line) and (15) (dahed
line) using the best fit values. As is evident, for z < 0.5 both functions
where we have assumed that both species have free Lagrangian are almost identical, but they start to differs from z > 0.5
of the dust type, i.e., with pm = 0 and px = 0 as equation of
state. From a variational point of view, as the one advocated
in the previous section, the stress energy density of the fluid
an expression that corresponds exactly to the LCDM flat
is obtained from
√ model. This means that it is enough to have two dust-like
δ( −gL) 2δL constituents in interaction, no matter how small they are (but
Tμν = −2 √ μν
= − μν + gμν L. (12)
−gδg δg non-zero), the resulting cosmological model is very similar to
If we assume that Lint does not contains kinetic terms, then a model without interaction between dust and a cosmological
δLint /δg μν = 0, and the stress energy tensor associated with constant.
μ μ Although we know this model is a very simple one, we
the interaction can be written as Tν = δν Lint , that inter-
preted in the context of a perfect fluid with energy density want to test it against observational data, to see to what extend
ρint = Lint , give rises naturally to a component with EoS the data adjust the values of the parameters (here α and β)
pint = −ρint . Using together this relation and Eq. (11) into away from the LCDM model.
Eq. (8) we get Because both contributions have the same form (see (14)),
there is no point in use two set of different parameters to
αρm0  
confront the data. Then we assume that
m =
x =
will
ρint = ρint
0
+ 1 − a −3(1+α)
1+α be the density parameter of both contributions, and that α =
βρx0   β = γ will be the interaction constant for both contributions
+ 1 − a −3(1+β) , (13)
1+β too. We know that this choice clearly select a very special
type of solution of the system, but we just want to explore
where ρint
0 = ρ (a = 1). From here is evident that as inter-
int how the model behaves once we constraint against the data.
actions turn off in this model, this implies that ρint
0 should
The best fit values obtained in this way are
= 0.11 ±
go to zero. In fact, from (13) replacing α = β = 0 leads to 0.03 and γ = 0.25 ± 0.15. Because the “effective” density
0 = 0. Of course, this happens only when the parameters
ρint parameter is the sum
m +
x = 2
then our model predicts
are exactly zero α = β = 0. a density parameter for dark matter of 2
= 0.22 ± 0.06. In
Notice that replacing (13) in (4) we get for E(z) = H/H0 Fig. 1 we display both (14) and (15) using the best fit values.
For small redshift, z < 0.5 both curves are almost identical,

m   but they differs appreciably for larger ones (Fig. 2).
E 2 (z) = 1 + (1 + z)3(1+α) − 1 +
1+α Let us discuss some points on ρint . Based on the solution

x   (13) and assuming that 1 + γ > 0 we get
+ (1 + z)3(1+β) − 1 , (14)
1+β
γ  0 
where we have defined
i = ρi0 /3H02 .At this point we ρint (z → −1) → ρint
0
+ ρm + ρx0 , (16)
1+γ
would like to point out the following. Suppose that the param-
eters α, β  1 are small (but not zero), then by expanding
so if we want to keep ρint ≥ 0 at any moment in the future,
(14) in series, we find that at zero order the Hubble function
we have to ensure that the right hand side of (16) must be
approaches:
≥ 0. For small values of α and β (or γ in this special case)
E 2 (z)  1 −
m −
x + (
m +
x )(1 + z)3 , (15) 0 + αρ 0 + βρ 0 ≥ 0.
this means that ρint m x

123
 862 Page 4 of 11 Eur. Phys. J. C (2020) 80:862

where
 
α 2 ρm0 (1 − a −3(1+α) )
F= − , (23)
β −α 1+α
and
 
β 2 ρm0 (1 − a −3(1+β) )
G= βρx0 + . (24)
β −α 1+β
Let us discuss this solution. Notice that as z → −1 (or
a → ∞) we get

α 2 ρm0
ρint → ρint
0
− (25)
(β − α) (1 + α)
 
β βρm0
+ ρx0 + . (26)
1+β β −α

Notice that within this limit and assuming small α, β  1

we can write

Fig. 2 Here we display the confidence contours – at one and two sigmas ρint → ρint
0
+ (α + β) ρm0 + ρx0 > ρint
0
, (27)
– for the two free parameters in the fit:
and γ defined in the text. Notice
that the data suggest a non zero γ at 1σ but at 2σ does not exclude the which certainly implies the condition (α + β)ρm0 + ρx0 ≥ 0,
zero value that seems easily to achieve.
On the other hand, the limit z → ∞ of (23) and (24) tells
Into the past, we would like to have ρint (z → ∞) > 0, us that
so from (13) we find that α 2 ρm0
F→ (1 + z)3(1+α) , (28)
αz 3(1+α) ρm0 βz 3(1+β) ρx0 (β − α) (1 + α)
ρint → ρint
0
− − , (17)  
1+α 1+β β βρm0
G→− ρx0 + (1 + z)3(1+β) , (29)
a result that seems difficult to achieve unless one of the 1+β β −α
parameters would be small but negative. and then

3.2 Asymmetric model α 2 ρm0

ρint → ρint
0
+ (1 + z)3(1+α) +
(β − α) (1 + α)
Let us assume the asymmetric interaction functions as  
β βρm0
− ρx0 + (1 + z)3(1+β) . (30)
Q m = −3H αρm , Q x = −3Hβ(ρm + ρx ) (18) 1+β β −α
Replacing the first one in Eq.(6) we obtain An expression that reduces to the following after considering
ρm (a) = ρm0 a −3(1+α) . (19) up to first order terms assuming α, β  1
 
From (7) and the second one in (18) we get ρint → ρint0
− β + α ρm0 + βρx0 (1 + z)3 , (31)

 −3(1+β)  which may implies that the term in square parenthesis should
a − a −3(1+α)
ρx (a) = ρx0 a −3(1+β) + βρm0 (20) be [...] ≤ 0 in order to get a ρint (z → ∞) ≥ 0. If we
β −α emphasize this point, we can express the condition even more
Following the previous case, we now write and solve the explicitly, (β + α) ρm0 + βρx0 = 0 or
equation for ρint . From (8)
ρm0 β
dρint =− . (32)
= 3H αρm + 3Hβ(ρm + ρx ). (21) ρx0 α+β
dt
Using the expressions (19) and (20) we find According to our best fit values for α and β both sides are
numbers close to one, and within the errors it is satisfied.
Notice that both limits points to consistent condition (see
ρint = ρint
0
+ F(a) + G(a), (22) (27) and (32)).

123
Eur. Phys. J. C (2020) 80:862 Page 5 of 11 862

Adding (19), (20) and (22) in (4) we find the Hubble func-
tion H (a). Evaluating it today we get
3H02 = ρm0 + ρx0 + ρint
0
, (33)
as it can be. Using this relation to replace ρint
0 , we can write

E = H/H0 and find

 
E 2 = 1 +
m −1 + a −3(1+α) + S(a) + R(a)
 
+
x −1 + a −3(1+β) + T (a) , (34)

where
a −3(1+β) − a −3(1+α)
S=β , (35)
β −α
β 2 1 − a −3(1+β) α 2 1 − a −3(1+α)
R= − , (36)
β −α 1+β β −α 1+α
1 − a −3(1+β)
T =β , (37)
1+β
In the limit for small α, and β we find at first order in the Fig. 3 Here we show the result of the statistical analysis using the
parameters that the function S(z) goes like Pantheon sample of type Ia supernovae

S(z)  −3β(1 + z)3 log(1 + z) + O(α 2 , αβ, β 2 ) (38)

as we have anticipated.
the R(z) function evolves as Further, as we have obtained above, for small α, β  1
 
R(z)  (α + β) 1 − (1 + z)3 + O(α 2 , αβ, β 2 ) (39) we get (38), (39) and (40), and then we can write up to first
order
and finally the T (z) behaves as   
  E 2 = 1 +
m [(α + β) − 1] 1 − (1 + z)3 +
T (z)  (α + β) 1 − (1 + z)3 + O(α 2 , αβ, β 2 ), (40)
− 3β (1 + z)3 log (1 + z)
this means that at order zero we get the same result as the  
previous section: the LCDM limit. +
x [(α + β) − 1] 1 − (1 + z)3 , (46)
Let us study the behavior of the solution in two limits: the
future at z → −1 and also into the distant past to z → ∞. in this way we get in the limit z → −1
Taking the expressions (35), (36), and (37) and looking for  3  
the limit z → −1 we get E 2 → 1 − (
m +
x ) − 3β
m 1 + z log 1 + z ,
β(1 + z)3  3β  3α (47)
S= 1+z − 1+z → 0, (41)
β −α

2  where the term in square brackets goes to zero as z → −1.
1 β α2 This means that independent of the value of β the CDM
R= − , (42)
β −α 1+β 1+α limit (for z → −1) is restored.
β It is then interesting to see what the analysis using obser-
T = (43)
1+β vational data can gives us about this model. Using the lat-
and if α, β << 1 which then implies that est supernova data [23] as we did in Sect. 3, we find the
following values as the best fit for the parameters:
m =



2 0.14 ± 0.09,
x = 0.146 ± 0.085, α = 0.09 ± 0.17,
1 β
E → 1 −
m 1 −
2
+ and β = −0.01 ± 0.17. The result of 5000 chains using
β −α 1+β
   the emcee code [21] is shown in Fig. 3). In order to test
α2 β
− −
x 1 − , (44) the convergence of our chains we follow the autocorrela-
1+α 1+β tion time. The integrated autocorrelation time quantifies the
Monte Carlo error and with it the efficiency of the sampler.
that after assuming α, β  1 reduces
Here we get τ f  100, so a run with 5000 chains are suffi-
E 2 (z → −1) → 1 − (
m +
x ) , (45) cient to be confident of the quality of our fit.

123
 862 Page 6 of 11 Eur. Phys. J. C (2020) 80:862

In order to check this, we need an observational probe

that constrain the model at large z as the cosmic background
radiation (CMB). Following [24] we use the three CMB dis-
tance priors, the shift parameter R, the acoustic scale la and
the baryon density parameter
b h 2 . The details of the com-
putation of the constraints from CMB follows [25] and is
explained in the Appendix, however is relevant to explain
certain points here. First of all, we have to add baryons
and radiation explicitly in the Friedman equation Eq. (4).
Because these two components conserved separately, Eq. (8)
is not modified. The best fit values obtained are α = 0.07,
β = −0.05 with
m = 0.142 and
x = 0.16. These values
Fig. 4 Here we display the deceleration parameter for our asymmet- were obtained in a joint analysis with SNIa and CMB pri-
ric model (dashed line) together with that from the flat LCDM model ors. A detailed study of this model at large redshift model
(continuous line) is underway, but we can conclude from these preliminary
results that: first, having inserted baryons and radiation, the
model is able to fit the data at large redshift with small inter-
action parameters in agreement with that implied by the SNIa
The data then implies small values for the parameters α data (at low redshift data). This means that our model is capa-
and β that controls the interaction between these dark matter ble to fit simultaneously both small and large redshift data
components. The small values of these values are in agree- keeping α, β small, and with a DM density
m +
x  0.3.
ment with our statement of getting close to the CDM model This conclusion is reinforced from a explicit computation of
in this limit. the age of the universe in this model. By using the Hubble
The total dark matter density parameter gives us
m + function H (z), with baryons and radiation added, and mak-

x  0.28 a value that is in agreement with other astrophys- ing use the best fit values for the parameters α, β,
m and
ical tests. From Fig. 3 it is clear the degeneracy between these
x , we get
two dark matter component, but keeping the sum essentially  ∞
dz
constant. t0 H0 =  0.9743, (48)
0 (1 + z)E(z)
Using these best fit values we can plot the reconstructed
deceleration parameter as a function of redshift z. We display which gives us an age of the Universe similar to that inferred
it in Fig. 4 together with the deceleration parameter for the from the LCDM model.
flat LCDM model with
m = 0.27. As we can see, both Another concern would be if the interaction functions here
curves essentially follows the same trend with a very small defined growth with redshift spoiling the matter formation
difference in amplitude (of our model compared to that from era. A simple way to check this, following [29], is consid-
the LCDM model). The redshift for the transition between ering the relative strength of the coupling f for each DM
deceleration to acceleration is around z  0.7 − 0.8 for the component: f m = Q m /3Hρm and f x = Q x /3Hρx . In the
models. case of the symmetric model we obtain f m = α and f x = β.
The best fit values obtained for the free parameters α and This also is obtained in the asymmetric model where f m = α
β are then small enough to obtain a model very close to the and f x  α. Because the strength parameters does not grow
LCDM at small redshift. It is then logic to ask for the perfor- with redshift, but reach constant small values, we can expect
mance of this model for large redshifts. To answer appropri- to obtain a similar behavior that those in [9] where a small
ately this question it necessarily take us beyond the scope of interaction parameters not only is possible, but it seems to
this work. However, we can try to answer some of the main be needed to obtain a better fit of the process of the structure
questions. In this case is imperative to add the contributions formation.
for both baryons and radiation that are negligible at small It is also interesting to see the evolution of each energy
redshift but which are increasingly important as we move to density component as a function of redshift. This is display
large z. However, we can not use the best fit values for α and in Fig. 5.
β obtained from the test using SNIa because these are valid Although ρint seems to keep a constant value, its value
for the range where data is available, this is 0.01 < z < 2.3. actually is decreasing with redshift until reaching a zero value
Necessarily the extension of our model beyond redshift 2 is at z  7 after which it becomes negative. However, we must
an extrapolation that can not be taken seriously. We need then be cautious with these results, since our physical system is
to test a modified model – with baryons and radiation – and of interaction between two components and therefore the
test it as we go back in time. analysis of each one separately, and especially the interaction

123
Eur. Phys. J. C (2020) 80:862 Page 7 of 11 862

let us comeback to the Eqs. (4)–(8) of Sect. 2. Defining the

total DM contribution ρ M = ρm + ρx , assuming both free
pressures as zero, the system of equations can be re-written
as

3H 2 = ρ M + ρint , (50)
Ḣ + 3H = − pint ,
2
(51)
ρ̇ M + 3Hρ M = Q , (52)
ρ̇int = −Q , (53)

where we have defined Q = Q m + Q x . These equations are

exactly those of the inhomogeneous vacuum model iVCDM
Fig. 5 Here we display each energy density component using the best proposed in [35]. In particular, and for the purpose of this
fit values obtained using type Ia supernovae + CMB priors constraints
paragraph, the linear perturbations of the model have been
also studied [36]. The idea is the following: to consider
presureless matter with p M = 0 and vacuum energy given
by ρV with EoS pV = −ρV such that

Tμν = TMμν + TV μν = ρ M u μ u ν − ρV gμν , (54)

where u μ is the matter 4-velocity. The stress-energy tensor

conservation reads then as

∇ μ TV μν = Q ν , ∇ μ Tmμν = −Q ν , (55)

where the interaction function is Q ν = −∇ν ρV . By choos-

ing Q μ = Qu μ , two things occurs, first the matter 4-velocity
is a potential flow and thus became irrotational, and second
Fig. 6 Here we plot the total equation of state for our model (continuos
the vacuum is homogeneous on hypersurfaces orthogonal to
line) in comparison to that of the LCDM model (dashed line) the matter 4-velocity, which means that there are not pressure
gradients in the frame comoving with matter, implying a mat-
ter sound speed equals to zero (as in LCDM). Although the
component, do not make much sense. Physically only the behavior of the perturbations may or may not take the exact
total density makes sense. form of the iVCDM model, it is essentially this similarity
We can also compute the total equation of state of our that can be considered a good indication even before calcu-
model using (4) and (5). Explicitly we just need to compute lating anything. We are working on this topic for a future
ptot ρint publication.
ωtot (z) = =− , (49)
ρtot ρm + ρx + ρint
which can be plotted using the best fit values recently
obtained. The results is display in Fig. 6 together with the 4 Thermodynamics
equivalent to the LCDM model. We notice a very small dif-
ference between them. Let us discuss the thermodynamics of the models proposed
As we see in the plot, at small redshifts, the effective EoS in previous section. We want to explore to what extend these
of our model is slightly larger than that of the LCDM model, new models for dark energy are consistent with the laws of
however this trends change around redshift z  2.1 after that thermodynamics.
the we f f of our model remains larger for a while until z  4
where again turns out to be larger. We have explore these 4.1 Symmetric case
changes for large redshift and always their difference (in the
total EoS parameter) is minor than 0.03. The fluid ρint behaves as  does in the sense that this contri-
Another point of interest is if this model allows for enough bution satisfy pint = −ρint . The fluid ρm behaves as one per-
ef f
growth of structure, given the interaction between these DM ceiving an effective small pressure pm = αρm and also the
ef f
components. Although this issue is beyond the scope of this fluid ρx as a fluid with px = βρx , where we have defined
ef f ef f
work, let us make some points on it. As a preliminary step the effective EoS parameters ωm = α and ωx = β. In

123
 862 Page 8 of 11 Eur. Phys. J. C (2020) 80:862

this way, the coupled system reduce to then its clear that d Sm (z → −1) → 0, and then also
d Sx (z → −1) → 0, so we restore the adiabaticitcy in this
ρ̇m + 3H (1 + α) ρm = 0, (56)
limit (the CDM limit). Notice that it seems relevant if
ρ̇x + 3H (1 + β) ρx = 0, (57) α > β for this conclusion to be correct. However, as we will
that can be studied along the discussion in [30] (see also [31]). discuss in brief, there is no real meaning to this inequality due
In this context, the temperature of the fluid can be written as to the symmetry of the model. Now if we assume α, β  1

 z  then ρm (z)  ρm0 (1 + z)3 , and ρx (z)  ρx0 (1 + z)3 , so we
dz e f f
T (z) = T 0 exp 3 ω (z) , (58) find
0 1+z

ρm0 Tx0 α
then we can write for each component d (Sm + Sx ) = 1 + 0 0 (1 + z) 6(α−β)
d Sx (72)
ρx Tm β
Tm (z) = Tm0 (1 + z)3α , Tx (z) = Tx0 (1 + z)3β , (59)
which implies there is no adiabaticity. So only when z → −1
in this way and according to [30] if α
= β
= 0 there is no it is possible to restore the adiabaticity. However, this is not
adiabaticity. This is recovered for α = β = 0. completely correct, because once we assume that α > β or
In general, without assuming an explicit EoS parameter α < β, we are making a choice about the future evolution
for the fluids, and according to the first law of the system. Further, since in this context
m and
x are
also interchangeable contributions, there is no need to worry
T d S = d[(+ p)V ] − V dp, (60)
about a particular hierarchy choice.
we can write for each component Let us discuss now the case for ρint . From the combination
d Sm d    dm of the first and second law
Tm = 1 + ωm m V − ωm V , (61)
dt dt dt Tint d Sint = d [(ρint + pint ) V ] − V dpint . (73)
d Sx d    dx
Tx = 1 + ωx x V − ωx V , (62)
dt dt dt Using the EoS for the compnent pint = −ρint we find
which can be rewritten as d Sint dρint
Tint =V , (74)
Tm d Sm dm dt dt
= + 3H (1 + ωm )m = Q m , (63)
V dt dt then, using the conservation equation for ρint we find
Tx d Sx dx
= + 3H (1 + ωx )x = Q x (64) Tint
V dt dt = − (Q m + Q x ) = 3H (αρm + βρx ) , (75)
where we have set d V /V dt = 3H . In this way V
Tm d Sm Tx d Sx from which we conclude
= Qm , = Qx (65)
V dt V dt d Sint
> 0, (76)
Tm Tx dt
⇒ d Sm = d Sx . (66)
Qm Qx and then, according to the relation,
So we can write


 Ṫint ∂ pint
Q m Tx = −3H = 3H, (77)
d (Sm + Sx ) = 1 + d Sx , (67) Tint ∂ρint
Q x Tm
and then if the temperature evolves as Tint (z) = Tint (0) (1 + z)−3 , so

 in the future limit Tint (z → −1) → ∞, a result typical of
α ρm Tx
d (Sm + Sx ) = 1 + d Sx
= 0, (68) models of DE.
β ρx Tm
We have to stress here that in the case of  we have p =
we know there is no adiabaticity. Replacing what we have −ρ = − and obviously ρ̇ = 0 so in this case we can not
found previously for temperatures we get use the relation Ṫ /T = −3H (∂ p/∂ρ). As we know [30]
ρm Tx ρ0 T 0 the temperature associated to  is zero, while in the present
= m0 m0 (1 + z)6(α−β) . (69) case pint = −ρint y ρ̇int
= 0, making evident an important
ρx Tm ρx Tx
difference between these two contributions.
In particular we observe that As we mentioned, the previous statement about the tem-
ρm Tx perature Tint (z → −1) → ∞ should not be a surprise for us.
(z → −1) → 0 (70) In fact, for a generic dark energy component pde = ωde ρde
ρx Tm
with ωde < 0 we have Ṫde /Tde = 3 |ωde | H, then the tem-
which implies that
perature evolves as Tde (z) = Tde (0) (1 + z)−3|ωde | , so in
d Sm (z → −1) + d Sx (z → −1) → d Sx (z → −1) , (71) the future limit we get Tde (z → −1) → ∞. In this sense

123
Eur. Phys. J. C (2020) 80:862 Page 9 of 11 862

ρint (z) plays a better role (a more physically stronger role) 5 Towards a field model
as dark energy than those played by .
It is interesting also to notice that in the context of CDM, In this section we describe a possible implementation of
the transition redshift between deceleration/acceleration the model presented in Sect. 3. For this, we will use two
occurs usually around z  0.5 which is also the redshift scalar fields whose free behavior, that is, ignoring interac-
from which our exact solution (14) start to differs from the tion between them, behave like dark matter, that is, a dust-like
CDM limit solution (15). evolution.
As a summary, we have a very simple model far more It is well known [32] that coherent scalar field oscillations
physically sound than the cosmological constant, where with a self interacting potential  φ n , behaves as a contri-
a very small coupling between dark matter components bution whose energy density decay as a −6n/(n+2) . For a pure
behaves as CDM. This is the case for the symmetric model, DM contribution, the energy density goes as a −3 then the
where the change α ⇐⇒ β and
m ⇐⇒
x left the Hubble potential would be V (φ) = V0 φ 2 . This is an exact result
function unchanged. In the next sub section we discuss the assuming that no other constituent than the scalar field is
asymmetric case. present.
Another way to build up a scalar field behaving as DM,
4.2 Asymmetric case is by using the reconstruction scheme. From [33] the scalar
field potential and kinetic term can be written in terms of the
Here we discuss the thermodynamics consequences of the scale factor through the parametric equations
asymmetric model previously presented. Let us start rewrit-

3 Ḣ ˙ 2 = − Ḣ .
ing the system of conservation equations for both compo- U (t) = H2 + , χ (t) (84)
8π G 3 4π G
nents. From (18) we have
Then by using a(t) = (t/t0 )2/3 typically of a dust like con-
ρ̇m + 3H (1 + α) ρm = 0, (78) tribution, we get H = 2/(3t) and Ḣ = −2/(3t 2 ), then from

 
ρm the field equation we get
ρ̇x + 3H 1 + β 1 + ρx = 0. (79)
ρx 1 √
χ̇ = √ , ⇒ ln t = 6π Gχ (t), (85)
ef f ef f
where ωm = α and ωx (z) = β [1 + ρm (z)/ρx (z)]. 6π Gt
Explicitly the quotient ρm /ρx takes the form then after we write the scalar field potential

 √
ρm ρ0 (1 + z)3(α−β) 3 2 e 6π Gχ
= m0   0 , (80) U (t) = ⇒ U (χ ) = , (86)
ρx ρx 1 + β ρm 8π G 9t 2 12π G
β−α ρ 0 1 − (1 + z) 3(α−β)
x
Then for the case of a dust evolution – a(t)  t 2/3 –
then assuming α > β we can take the future limit the equations leads us to an exponential potential U (χ ) =
√
z −→ −1 we find that ρm /ρx → 0 which implies that U0 exp(−αχ ), in which α = 6π G. This is a well know
ef f
ωx (z → −1) → β. On the other hand, in the limit of the result [34].
ef f ef f
far past z −→ ∞ we get that ωx (z → ∞) → α = ωm . In our model then, we consider these two scalar fields φ(t)
Now, let us compute explicitly the temperatures. Using and χ (t) interacting through
the formula (58) this leads to
1 2 1 1
L= φ̇ − V (φ) + χ̇ 2 − U (χ ) − g 2 φ 2 χ 2 . (87)
Tm = Tm0 (1 + z) 3α
, (81) 2 2 2

 
z dz ρm (z) The stress energy tensor for the homogeneous free fields φ(t)
Tx = Tx (1 + z) exp 3β
0 3β
(82) and χ (t) can be written as those of a perfect fluid with energy
0 1 + z ρx (z)
density and pressure given by
Following the same steps described in the analysis of the
1 2 1
symmetrical case, we have ρφ = φ̇ + V (φ), pφ = φ̇ 2 − V (φ) (88)
  2 2
Q m Tx The interaction Lagrangian can also be written in a perfect
d (Sm + Sx ) = 1 + d Sx ,
Q x Tm fluid form, but this time the energy density and pressure are

 
α ρm 1 Tx
= 1+ d Sx , (83) ρφχ =
1 2 2 2
g φ χ ,
1
pφχ = − g 2 φ 2 χ 2 ,
β ρx 1 + ρm /ρx Tm 2 2
(89)

which implies no-adiabaticity. However, as we take the limit which – as we have anticipated – automatically satisfies the
z → −1, then ρm /ρx → 0, then we obtain that both d Sm → cosmological constant equation of state, although the energy
0 and d Sx → 0, and the adiabaticity is restore in this limit. density is not constant.

123
 862 Page 10 of 11 Eur. Phys. J. C (2020) 80:862

The field equations are: the Friedman equation (4) and We have also performed a statistical analysis using the lat-
(5) with pressures and densities defined by (88) and (89) est data set for type Ia supernova (the Pantheon sample [23])
before (where the notation ρint = ρφχ ), and the well known consisting in 1048 data points and its covariance. Although
very simplistic – because we have not added a explicit cur-
φ̈ + 3H φ̇ + V  (φ) = −g 2 χ 2 φ, (90)
vature or a baryonic term or radiation term in the Hubble
and function – our models are able to describe successfully the
data, with small best fit values for the parameters, being in
χ̈ + 3H χ̇ + U  (χ ) = −g 2 φ 2 χ , (91)
agreement with the hypothesis of the model. Although the
which are the equivalent to (6) and (7), where we can identify errors are big, the contrast with observational data implies
the existence of a large family of models with small α and β
Q m = −g 2 χ 2 φ φ̇, Q x = −g 2 φ 2 χ χ̇. (92)
parameters, that describe an evolution that mimic the CDM
From (92) and (89) is clear that (8) is automatically satisfied. model without the necessity to add an exotic dark component.
This model clearly show the way we can built a field model
of two DM component in interaction with an evolution sim- Acknowledgements We are grateful to Miguel Angel Cruz for helpful
discussions.
ilar to the CDM model. A work in progress in underway
where we focus on this specific model. Data Availability Statement This manuscript has no associated data
or the data will not be deposited. [Authors’ comment: The data used in
this paper has been already published in Ref. [23].]
6 Discussion Open Access This article is licensed under a Creative Commons Attri-
bution 4.0 International License, which permits use, sharing, adaptation,
In this paper we have proposed a family of models for DE distribution and reproduction in any medium or format, as long as you
consisting in two DM species interacting each other, whose give appropriate credit to the original author(s) and the source, pro-
vide a link to the Creative Commons licence, and indicate if changes
interaction although small, enable us to describe a typical were made. The images or other third party material in this article
evolution of the CDM model. The key element is the role are included in the article’s Creative Commons licence, unless indi-
accomplished by the energy density associated to the interac- cated otherwise in a credit line to the material. If material is not
tion Lagrangian, ρint . Assuming the interaction Lagrangian included in the article’s Creative Commons licence and your intended
use is not permitted by statutory regulation or exceeds the permit-
does not have derivative couplings, the automatic equation of ted use, you will need to obtain permission directly from the copy-
state that this component satisfy is pint = −ρint i.e., that of right holder. To view a copy of this licence, visit http://creativecomm
the cosmological constant. However, although this compo- ons.org/licenses/by/4.0/.
nent satisfy this EoS, the energy density evolves (in contrast Funded by SCOAP3 .
to  that keep its value constant), making it a more sound
component physically speaking. For example, this does not
suffer from the “coincidence problem” because the interac- Appendix
tion energy density – which is interpreted here as the equiv-
alent to  – emerges from a Lagrangian that connects both Here we describe the formulae to use CMB priors to con-
DM species from the beginning. This connection also answer straint our model. This analysis follows [24] and [25]. We
our question about the order of magnitude of . Here the use CMB information by using the Planck data [26] extracted
response is in essence because both DM contribution are tied from the analysis performed by [24] to probe expansion his-
through the interaction which established the order of mag- tory up to the last scattering surface. The χ 2 for the CMB
nitude of their contributions. Although obvious, it is also data is constructed as
necessary to highlight the fact that we do not need an exotic
(negative pressure) component to describe the observations. χC2 M B = X T CC−1M B X, (A.1)
In fact, from a thermodynamic point of view, this com-
ponent behaves more naturally than , showing a tempera- where, for a flat universe the data vector to consider is
ture that increases in the future, a behavior typical to other (R, l A ,
b h 2 ) with
DE models where the EoS parameter varies with redshift, ⎛ ⎞
in contrast to the  behavior where the temperature associ- 1.74963
ated is zero. Furthermore, we have discussed how the non- X = ⎝ 301.80845 ⎠ . (A.2)
adiabaticity emerges from the model, clearly due to the mani- 0.02237
fest interaction, and its future evolution towards adiabaticity.
Here l A is the “acoustic scale” defined as
Certainly, a much more physical behavior than the discon-
nected evolution between  and the rest of the constituents π d L (z ∗ )
of the universe that is evident in the CDM model. lA = , (A.3)
(1 + z)rs (z ∗ )

123
Eur. Phys. J. C (2020) 80:862 Page 11 of 11 862

where d L (z) is the proper luminosity distance and the redshift 14. M. Klasen, M. Pohl and G. Sigl, Prog. Part. Nucl. Phys.
of decoupling z ∗ is given by [27], 85 (2015), 1–32 https://doi.org/10.1016/j.ppnp.2015.07.001
arXiv:1507.03800 [hep-ph]]
  −0.738    g2 15. C. Boehm, J.A. Schewtschenko, R.J. Wilkinson, C.M. Baugh, S.
z ∗ = 1048 1 + 0.00124
b h 2 1 + g1
m h 2 ,(A.4) Pascoli, Mon. Not. Roy. Astron. Soc. 445, L31 (2014). https://doi.
org/10.1093/mnrasl/slu115. arXiv:1404.7012 [astro-ph.CO]
 −0.238 16. J. A. Schewtschenko, R. J. Wilkinson, C. M. Baugh, C. Bœhm
0.0783
b h 2 0.560
g1 =  0.763 , g2 =  1.81 , and S. Pascoli, Mon. Not. Roy. Astron. Soc. 449, no. 4, 3587
1 + 39.5
b h 2 1 + 21.1
b h 2 (2015) https://doi.org/10.1093/mnras/stv431 arXiv:1412.4905
(A.5) [astro-ph.CO]
17. J. A. Schewtschenko, C. M. Baugh, R. J. Wilkinson, C. Bœhm,
S. Pascoli and T. Sawala, Mon. Not. Roy. Astron. Soc.
The “shift parameter” R defined as [28] 461, no. 3, 2282 (2016). https://doi.org/10.1093/mnras/stw1078
√ arXiv:1512.06774 [astro-ph.CO]

m 18. R.J. Wilkinson, J. Lesgourgues, C. Boehm, JCAP 1404,
R= D L (z). (A.6)
c(1 + z ∗ ) 026 (2014). https://doi.org/10.1088/1475-7516/2014/04/026.
arXiv:1309.7588 [astro-ph.CO]
CC−1M B in Eq. (A.1) is the inverse covariance matrix, 19. A. D. Dolgov, S. L. Dubovsky, G. I. Rubtsov and I. I. Tkachev,
Phys. Rev. D 88, no. 11, 117701 (2013). https://doi.org/10.1103/
⎛ ⎞ PhysRevD.88.117701 arXiv:1310.2376 [hep-ph]
1598.9554 17112.007 −36.311179 20. N. Tamanini, Phys. Rev. D 92, no. 4, 043524 (2015)
CC−1M B = 10−8 ⎝ 17112.007 811208.45 −494.79813 ⎠ . arXiv:1504.07397 [gr-qc]
−36.311179 −494.79813 2.1242182 21. D. Foreman-Mackey, D.W. Hogg, D. Lang et al., PASP 125, 306
(A.7) (2013)
22. J. Goodman, J. Weare, Commun. Appl. Math. Comput. Sci. 5, 65
(2010)
23. D. Scolnic, et al. Astrophys. J. 859 (2018) no.2, 101
arXiv:1710.00845 [astro-ph.CO]
24. Z. Zhai, Y. Wang, JCAP 07, 005 (2019). arXiv:1811.07425 [astro-
References ph.CO]
25. V.H. Cárdenas, Phys. Lett. B 750, 128 (2015)
1. A.G. Riess et al., Astron. J. 116, 1009 (1998) 26. N. Aghanim et al. [Planck], arXiv:1807.06209 [astro-ph.CO]
2. S. Perlmutter et al., Astrophys. J. 517, 565 (1999) 27. W. Hu, N. Sugiyama, ApJ 471, 542 (1996)
3. J. Frieman, M. Turner, D. Huterer, Ann. Rev. Astron. Astrophys. 28. J.R. Bond, G. Efstathiou, M. Tegmark, MNRAS 291, L33 (1997)
46, 385 (2008) 29. C.G. Böhmer, G. Caldera-Cabral, N. Chan et al., Phys. Rev. D 81,
4. J. Martin, Mod. Phys. Lett. A 23, 1252 (2008). arXiv:0803.4076 083003 (2010)
[astro-ph] 30. V. H. Cárdenas, D. Grandón and S. Lepe, Eur. Phys. J. C79 (2019)
5. S. Tsujikawa, Lect. Notes Phys. 800, 99 (2010). arXiv:1101.0191 no.4, 357, arXiv:1812.03540
[gr-qc] 31. R. Maartens, arXiv:astro-ph/9609119
6. M. Kunz, S. Nesseris and I. Sawicki, Phys. Rev. D 92, no. 32. M.S. Turner, Phys. Rev. D 28(6), 1243 (1983)
6, 063006 (2015). https://doi.org/10.1103/PhysRevD.92.063006 33. T. Padmanabhan, Phys. Rev. D 66, 021301 (2002), see also section
arXiv:1507.01486 [astro-ph.CO] 4.3 of T. Padmanabhan, Theoretical Astrophysics, Vol III: Galaxies
7. B. Wang, E. Abdalla, F. Atrio-Barandela and D. Pavon, Rept. Prog. and Cosmology, Cambridge University Press, Cambridge (2002)
Phys. 79, no. 9, 096901 (2016) arXiv:1603.08299 [astro-ph.CO] 34. A.R. Liddle, R.J. Scherrer, Phys. Rev. D 59, 023509 (1999). https://
8. R. Murgia, S. Gariazzo and N. Fornengo, JCAP 1604, no. 04, 014 doi.org/10.1103/PhysRevD.59.023509. arXiv:astro-ph/9809272
(2016) arXiv:1602.01765 [astro-ph.CO] [astro-ph]
9. V. Salvatelli, N. Said, M. Bruni, A. Melchiorri and D. Wands, Phys. 35. D. Wands, J. De-Santiago, Y. Wang, Class. Quant. Grav.
Rev. Lett. 113, no. 18, 181301 (2014) arXiv:1406.7297 [astro- 29, 145017 (2012). https://doi.org/10.1088/0264-9381/29/14/
ph.CO] 145017. arXiv:1203.6776 [astro-ph.CO]
10. A. Pourtsidou, C. Skordis and E. J. Copeland, Phys. Rev. D 88, no. 36. H. A. Borges and D. Wands, Phys. Rev. D 101 (2020) no.10, 103519
8, 083505 (2013). arXiv:1307.0458 [astro-ph.CO] https://doi.org/10.1103/PhysRevD.101.103519 arXiv:1709.08933
11. H. Velten, T. R. P. Caramês, J. C. Fabris, L. Casarini and [astro-ph.CO]
R. C. Batista, Phys. Rev. D 90, no. 12, 123526 (2014).
arXiv:1410.3066 [astro-ph.CO]
12. E. Ferreira, G.M., J. Quintin, A. A. Costa, E. Abdalla and B. Wang,
Phys. Rev. D 95 (2017). no.4, 043520 arXiv:1412.2777 [astro-
ph.CO]
13. R. Erdem, Phys. Dark Univ. 15, 57 (2017). arXiv:1612.04864 [gr-
qc]

123